🧬 Aging Wiki

multivitamins

Properties1
aliasesmultivitamin, multivitamin-mineral supplement, MVM, MVMS, daily multivitamin, one-a-day

9 min read

Multivitamin-mineral supplements

A multivitamin-mineral supplement is a single dosage form combining most essential vitamins and several trace minerals at or near nutritional (Daily-Value-scale) doses. It is the most widely consumed dietary supplement in the developed world. As an aging intervention it is best understood not as a geroprotector with a mechanism of action but as insurance against marginal micronutrient deficiency — which frames both what the evidence shows and why it shows it.

The high-quality evidence converges on a consistent conclusion: in well-nourished populations, daily multivitamin use does not reduce all-cause mortality, cardiovascular disease, or total cancer. The single replicated positive signal is on cognition in older adults. Formulation matters more than the “multivitamin” label suggests — the few ingredients with genuine harm signals (beta-carotene, high-dose vitamin E, iron in iron-replete men) are present in some products and absent in others.

Hard-endpoint evidence (mortality, cancer, cardiovascular disease)

StudyDesignnEndpointResult
Loftfield 2024 1Pooled prospective cohort (3 US cohorts), ≤27 yr390,124All-cause mortalityNull — HR 1.04 (1.02–1.07) first 12 yr; 1.04 (0.99–1.08) thereafter. Cause-specific (cancer/heart/cerebrovascular) all null.
COSMOS (Sesso 2022) 2Randomized, double-blind, placebo-controlled, women ≥65/men ≥6021,442Total cancer; cardiovascular diseaseNull — no significant reduction in total invasive cancer (HR 0.97, 95% CI 0.86–1.09) or total CVD (HR 0.98, 95% CI 0.86–1.12)
Physicians’ Health Study II (Gaziano 2012) 3Randomized, double-blind, placebo-controlled, male physicians, 11.2 yr median14,641Total cancerModest reduction — HR 0.92 (0.86–0.998), p=0.04; no cardiovascular or mortality benefit. Older formulation (Centrum Silver, 1997-era), older men.

Interpretation. The randomized trials (which break the confounding that plagues cohort data) and the largest cohort analysis agree: no effect on death, and at best a small, borderline cancer signal in one older male cohort that has not been reproduced. The 2022 US Preventive Services Task Force review reached the same place — insufficient evidence to recommend multivitamins for cardiovascular or cancer prevention, with an explicit recommendation against beta-carotene (raises lung-cancer risk in at-risk groups) and vitamin E (no benefit) 4.

DimensionStatus
Mechanism plausible in humans?partial — repletion corrects deficiency, but most trial populations were already replete
Hard-endpoint benefit in humans?no (mortality, cardiovascular disease, total cancer)
Replicated?yes — the null is replicated across cohort + multiple RCTs

The cognition exception

The one place a clean multivitamin has shown a replicated randomized benefit is cognition in older adults, via sub-studies of COSMOS:

  • COSMOS-Web (Yeung 2023): n=3,562, web-based memory testing; the multivitamin arm improved immediate recall significantly at 1 year (p=0.025), sustained over 3 years (p=0.011) — estimated equivalent of 3.1 years of age-related memory change 5.
  • COSMOS-Mind (Baker 2022): n≈2,260, telephone cognitive battery; multivitamin slowed global cognitive aging by an estimated ~60% (≈1.8 years) over 3 years 6.
  • A pooled meta-analysis of the three COSMOS cognitive studies confirmed a consistent modest benefit 7.

Caveats. The effect sizes are small, the mechanism is unestablished (possibly correction of subclinical B-vitamin or other deficiency in elders with declining intake/absorption), and the populations were ≥60 years old (mean ~73). Whether the benefit extends to cognitively healthy midlife adults is untested — this is the live evidence gap (see next-experiment).

Why cohort mortality signals mislead: healthy-user / sick-user confounding

Observational multivitamin–mortality associations are dominated by confounding that runs in both directions:

  • Healthy-user bias — habitual supplement users exercise more, smoke less, and eat better, biasing toward apparent benefit.
  • Sick-user / reverse causation — people often start a multivitamin when health declines (“I should take better care of myself”), biasing toward apparent harm.

The slight HR>1 in Loftfield’s first follow-up period is almost certainly residual sick-user confounding, not toxicity; the authors decline to claim harm 1. This is the central reason a cohort mortality null (or slight positive) cannot be read as “multivitamins are useless or dangerous” — the design cannot resolve a near-zero true effect from confounding of comparable magnitude.

Formulation matters — the harm subset

“Multivitamin” is not one product. Across the cohorts and legacy formulations, the pill frequently contained three ingredients with real harm or null signals, which a modern clean formulation can omit:

  • Beta-carotene — increases lung-cancer risk in smokers and asbestos-exposed individuals (ATBC, CARET trials); USPSTF recommends against it 4.
  • High-dose vitamin E (≥400 IU) — increased prostate-cancer incidence in the SELECT trial; no cardiovascular or cancer benefit. Nutritional doses (~10–15 IU / 7–11 mg) carry no such signal.
  • Iron — appropriate in menstruating women and the iron-deficient, but a pro-oxidant cardiovascular and iron-overload concern in iron-replete adult men, for whom an iron-free formulation is preferred.

A product that omits all three (iron-free, beta-carotene-free, nutritional-dose vitamin E) has a near-zero downside profile, which is the relevant safety frame for an individual decision even when the population benefit is null.

Who plausibly benefits

Because the mechanism is repletion, expected value concentrates where a deficiency gap exists:

  • Marginal or overt micronutrient deficiency (restricted/low-calorie diets, intermittent fasting or skipped meals, malabsorption, bariatric surgery, older adults with declining intake/absorption, certain medications).
  • Specific genotype/intake mismatches — e.g., a reduced-function MTHFR variant paired with low folate intake, where the methylated folate (5-MTHF) form is the appropriate choice over folic acid.
  • Populations with documented low diet quality (a plausible source of the COSMOS cognition effect).

In a well-measured, replete individual, the marginal expected value approaches zero — there is little deficiency gap to fill, so neither benefit nor harm should be expected from a clean formulation. The decision in that setting is genuinely close to neutral and turns on cost, simplicity preference, and risk tolerance rather than on outcome data.

Note on homocysteine-lowering B-vitamins

Multivitamins supply folate, B12, and B6, which lower plasma homocysteine. The inference that this should reduce cardiovascular events has been tested and largely failed: randomized B-vitamin homocysteine-lowering trials (HOPE-2, SEARCH, VITATOPS) did not reduce cardiac events, with at most a modest stroke-risk reduction in some analyses 8. So even the most mechanistically coherent multivitamin component lacks hard-outcome support for cardiovascular prevention — relevant when weighing the B-vitamin/methyl-donor rationale for any individual.

Recent synthesis (2026)

A 2026 umbrella review of 19 meta-analyses (~5.5 million participants) is consistent with the framing above on the two main endpoints — multivitamin use was associated with better global cognition, episodic memory, and immediate recall in older adults, and with no mortality benefit — but it additionally reported an association between multivitamin use and higher risk of age-related macular degeneration (AMD) progression 9. This last signal is unexpected (antioxidant-plus-zinc AREDS-type formulations are used clinically to slow AMD progression) and most likely reflects confounding by indication or formulation heterogeneity across the pooled meta-analyses rather than a causal harm; it has not been demonstrated in a randomized multivitamin trial. Flagged here for completeness pending a formulation-resolved analysis. contradictory-evidence

Footnotes

  1. Loftfield E et al. “Multivitamin Use and Mortality Risk in 3 Prospective US Cohorts.” JAMA Network Open 2024 · doi:10.1001/jamanetworkopen.2024.18729 · n=390,124 (NIH-AARP 327,732; PLCO 42,732; AHS 19,660) · pooled prospective cohort · follow-up up to 24 yr (NIH-AARP), 27 yr (PLCO), 26 yr (AHS) · all-cause mortality multivariable HR 1.04 (95% CI 1.02–1.07) FP1; 1.04 (95% CI 0.99–1.08) FP2; cause-specific HRs (heart disease, cancer, cerebrovascular) all null · authors attribute HR>1 to residual confounding, decline to claim harm · model: US adults free of cancer and major chronic disease at baseline ↩ ↩2

  2. Sesso HD et al. “Multivitamins in the prevention of cancer and cardiovascular disease: the COSMOS randomized clinical trial.” Am J Clin Nutr 2022 · doi:10.1093/ajcn/nqac056 · n=21,442 · randomized · total cancer HR 0.97 (95% CI 0.86–1.09) p=0.57; total CVD HR 0.98 (95% CI 0.86–1.12) — both null · model: US adults (women ≥65, men ≥60; mean age 72.1) ↩

  3. Gaziano JM et al. “Multivitamins in the Prevention of Cancer in Men: the Physicians’ Health Study II RCT.” JAMA 2012 · doi:10.1001/jama.2012.14641 · n=14,641 · randomized · total cancer HR 0.92 (0.86–0.998) p=0.04; no cardiovascular/mortality benefit · model: male US physicians ≥50 yr ↩

  4. US Preventive Services Task Force (Mangione CM et al.). “Vitamin, Mineral, and Multivitamin Supplementation to Prevent Cardiovascular Disease and Cancer: US Preventive Services Task Force Recommendation Statement.” JAMA 2022 · doi:10.1001/jama.2022.8970 · PMID 35727271 · evidence review · I statement (insufficient evidence) for multivitamins; D recommendation (recommends against) beta-carotene (harms outweigh benefits with moderate certainty) and vitamin E (no net benefit with moderate certainty) for CVD/cancer prevention ↩ ↩2

  5. Yeung LK et al. “Multivitamin Supplementation Improves Memory in Older Adults: A Randomized Clinical Trial (COSMOS-Web).” Am J Clin Nutr 2023 · doi:10.1016/j.ajcnut.2023.05.011 · n=3,562 (ITT) · randomized · improved immediate recall at 1 yr (p=0.025), sustained over 3 yr (p=0.011); effect ≈3.1 yr of age-related memory change · model: adults ≥60 yr (mean 71.0) ↩

  6. Baker LD et al. “Effects of cocoa extract and a multivitamin on cognitive function: a randomized clinical trial (COSMOS-Mind).” Alzheimer’s & Dementia 2023 · doi:10.1002/alz.12767 · n=2,262 · randomized · global cognition benefit: mean z=0.07 (95% CI 0.02–0.12) p=0.007 over 3 yr; estimated ~1.8 yr (60%) slowing of cognitive aging · model: adults (women ≥65, men ≥60; mean age 73, 60% women) ↩

  7. Vyas CM et al. “Effect of multivitamin-mineral supplementation versus placebo on cognitive function: results from the COSMOS clinic subcohort and meta-analysis of 3 COSMOS cognitive studies.” Am J Clin Nutr 2024 · doi:10.1016/j.ajcnut.2023.12.011 · PMID 38244989 · meta-analysis (COSMOS-Clinic n=573, COSMOS-Mind n=2,158, COSMOS-Web n=2,472; nonoverlapping) · global cognition benefit: mean difference 0.07 SD (95% CI 0.03–0.11) p=0.0009; episodic memory 0.06 SD (95% CI 0.03–0.10) p=0.0007; equivalent to ~2 yr reduction in cognitive aging · model: COSMOS participants ≥60 yr ↩

  8. Representative homocysteine-lowering RCT: Lonn E et al. (HOPE-2 Investigators). “Homocysteine Lowering with Folic Acid and B Vitamins in Vascular Disease.” N Engl J Med 2006 · doi:10.1056/NEJMoa060900 · PMID 16531613 · n=5,522 · randomized · primary composite (CV death, MI, stroke) RR 0.95 (95% CI 0.84–1.07) p=0.41 — null; stroke RR 0.75 (95% CI 0.59–0.97) p=0.03 — modest reduction · model: adults ≥55 yr with vascular disease/diabetes. Concordant findings reported in SEARCH and VITATOPS. contradictory-evidence (mechanism plausible, hard endpoints null) ↩

  9. Wang Y et al. “Umbrella review of multivitamin/mineral supplementation and health outcomes.” Ageing Research Reviews 2026 · 114:102965 · PMID 41308839 · umbrella review of 19 meta-analyses, n≈5,535,426 · multivitamin associated with improved global cognition / episodic memory / immediate recall in older adults; no all-cause mortality benefit; reported association with higher AMD-progression risk (likely confounded by indication/formulation; not from a randomized multivitamin trial) · model: pooled (predominantly older-adult) populations. Surfaced via 2026-06-02 supersession check; AMD signal not independently verified against the included primary studies. ↩

Graph View